The current issue and full text archive of this journal is available at www.emeraldinsight.com/1463-5771.

htm

Benchmarking procurement functions: causes for superior performance

Rupert A. Brandmeier
Transformation Consulting International, Mannheim, Germany, and

Benchmarking procurement functions 5

Florian Rupp
Fakultat fur Mathematik, Technische Universitat Munchen, Garching, Germany
Abstract
Purpose The purpose of this paper is to gain a better understanding between procurement strategy, organization, processes, methods and tools, human resources, supplier management, and overall procurement success. Design/methodology/approach In order to achieve the above-mentioned results, a holistic benchmark with highly recognized companies is conducted. Applying a cohesive questionnaire of open and closed questions, the paper covers all relevant aspects of procurement functions. Regression analysis is used to identify signicant correlations. Findings The benchmarking (BM) proved the following signicant success factors for the overall procurement process: use of cross-functional teams, high hierarchical positioning of the procurement function within the company, strong cooperation with other functions, training and development of the procurement personnel as well as supplier integration and continuous evaluation. Practical implications The paper provides the reader with sound evidence of how to improve the overall performance of procurement. Originality/value The analytical results of the research rely on statistical/mathematical methodology to substantiate qualitative BM results. Keywords Benchmarking, Supply chain management, Procurement, Regression analysis, Correlation analysis Paper type Research paper

1. Introduction Roughly over 60 percent of a companys spend amounts to procurement/supply chain management (SCM) expenses[1]. Especially, in competitive sectors and during recent crisis the strategic signicance of this function cannot be denied, and lots of efforts are continuously put into practice to strengthen procurement units. In particular, benchmarking (BM) of operative and strategic tasks served as a powerful method to identify weak spots as well as procurement best practices, see, e.g. Youssef and Zairi (1996a), Le Sueur and Dale (1997), Homburg et al. (1997), Gilmour (1998), Andersen et al.
The authors would like to thank the anonymous referee for her/his suggestions to more stringently display the results of this paper and especially for her/his speed when delivering feedback. It is awesome to submit a paper at Friday noon and receive the profound referees report the following Monday morning.
Benchmarking: An International Journal Vol. 17 No. 1, 2010 pp. 5-26 q Emerald Group Publishing Limited 1463-5771 DOI 10.1108/14635771011022299

BIJ 17,1

(1990), Frehner and Bodmer (2000), Sanchez-Rodrguez et al. (2003), A.T. Kearney (2004), Aberdeen Group (2006), Saad and Patel (2006), Schuh et al. (2007), Wong and Wong (2008) or Raymond (2008). Often, time demands require a careful selection of approximately ve BM partners with whom actual interviews are conducted a number to small to really apply powerful statistical tools. The alternatives often are broad surveys of hundreds of companies where just some more or less specic questions are issued. Here, some middle way is discussed which allows the application of statistical methods such as regression analysis, to analyze causes and effects of best practice. See also Codling (1997) and Codling (1998) for a multidimensional analysis of benchmark ndings and their incorporation into a company. We had been fortunate to execute a rather extensive individual BM project for an international automotive company, where, after a careful selection of possible cross industry BM partners, 14 highly recognized companies agreed to be interviewed (Brandmeier et al., 2008). The assessment type questionnaire guiding these interviews contained over 170 closed and open questions. This allowed us to apply evolved statistical methods and achieve sound cause and effect statements for the set-up of a tangible business plan. Thus, we intend to provide an extensive look into the data processing processes and how correlations and insights can be distilled from the information gathered by BM interviews. 2. Research background and literature survey Considering BM as the method of choice to determine procurement best practice is, for instance, conrmed by Sanchez-Rodrguez et al. (2003): this evaluation of over 300 companies shows a signicant positive impact of BM on purchasing performance and an indirect positive effect on business performance. Raymond (2008) comes to the same conclusion with respect to public instead of private procurement. Moreover, Wong and Wong (2008) present a detailed literature survey on 16 research articles dealing with aspects of procurement best practice dating from 1995 to 2003. They also highly emphasize the requirement to analyze data from supply chain benchmarks in a rigorous way. The consequent next step is to ask, what the practices are that constitute a superior performing procurement unit and to validate these ndings mathematically. Empirical studies show a huge bunch of different sometimes interwoven, sometimes clearly separated key features/levers that need to be strengthened in order to gain overall procurement success. Table I classies some approaches to drive procurement to superior performance. Hereby, we classied the indicators given in the selected sources (Frehner and Bodmer, 2000; A.T. Kearney, 2004; Aberdeen Group, 2006; Schuh et al., 2007) according to six elds of excellence with cover all activities of a procurement function. At a rst glance it is rather obvious, that every aspect counts. Following the combined guidelines of the research given in Table I and the previously discussed sources the procurement function needs to be perfect in any task operated and simultaneously guided by the most accurate strategy. Thus, it is reasonable to come back to the ground and ask what the really essential aspects of procurement success are. We do not need to measure everything that matters; we only need to measure the things that matter (Saad et al., 2005, pp. 383-97). Hence, are there a few key features

Field of excellence Strategy

Indicatora

Sources

Organization Processes Methods and tools

HRs

Supplier management

Precisely dened and communicated Frehner and Bodmer (2000) strategy Senior management support for procurement P as driver for company-wide saving activities Early involvement of P in development projects Right key performance indices Frehner and Bodmer (2000) and Aberdeen Group (2006) Early involvement of key suppliers in A.T. Kearney (2004) development projects Advanced cost cutting methods/levers (A.T. Kearney (2008), too) Risk management w.r.t. to future A.T. Kearney (2004), Aberdeen evolution possibilities of suppliers Group (2006) and Schuh et al. (2007) Corporate thinking and cross-functional Aberdeen Group (2006) and responsibility for all expenses Schuh et al. (2007) Global sourcing w.r.t. total cost of Schuh et al. (2007) ownership Central coordination and local execution Schuh et al. (2007) Standardized procurement processes Frehner and Bodmer (2000) Procurement handbook Frehner and Bodmer (2000) Intranet as procurement knowledge base Continuous establishment of data transparency e-Procurement Shared e-platform with suppliers Methods for forecasting, inventory Aberdeen Group (2006) management, and replenishment Highly qualied buyers Frehner and Bodmer (2000) P personnel must be on face value with members of other units (as development, production, etc.) Specialized procurement roles Schuh et al. (2007) Structured supplier portfolio Frehner and Bodmer (2000), A.T. Kearney (2004) and Schuh et al. (2007) Holistic supplier evaluation Frehner and Bodmer (2000) and Schuh et al. (2007) Cost reduction by supplier development Schuh et al. (2007) Supplier value integration Management of sub-suppliers

Benchmarking procurement functions 7

Note: aAs provided in source; here P denotes procurement Sources: Frehner and Bodmer (2000), A.T. Kearney (2004), Aberdeen Group (2006) and Schuh et al. (2007)

Table I. Success factors for the procurement function (P) proposed in the recent sources/surveys

BIJ 17,1

of superior performance that should be known by any manger? Well, let us benchmark companies acknowledged for their procurement performance[2], gather the raw data and analyze them. 3. Research methodology The overall BM process conducted does not deviate from that of other studies, see Camp (1989) and, in particular, Brandmeier et al. (2008). That is to say, rst the key issues where dened and the questionnaire developed and tested by company internal interviews[3] (Carpinetti and de Melo, 2002). Then, best practice companies were identied, selected and contacted; cf. Razmi et al. (2000) for guidelines for the identication/selection process of best practice BM partners. Third, the interviews were conducted at the locations of the BM partner. Usually, an interview took three to ve hours whereby one to two members of the partner were interviewed by two members of the BM team (one asked the questions issued in the questionnaire and the other took notes on the answers[4]). Finally, the interesting work begins: the evaluation of the data received via the interviews. Here, a brief overview is provided on the questions/themes issued during the BM interviews, together with an evaluation methodology to end up with an unbiased numerical classication of the answers. Last, we dene the excellence of a procurement function by means of their handling of 12 procurement levers. This setting will provide the ground for the mathematical analysis in Section 4. 3.1 The questionnaires internal structure Yasin (2002) already signied that the direction of addressing BM especially in the complex world of supply chains and procurement is no longer process oriented, but rather on an holistic approach encompassing strategies and systems orientation. This is reected by our questionnaire, which deals with the following six assessment elds (Section 2). Each of these elds is being divided into ner clusters to provide a more integrated view on the subject (the titles of the sub-elds are given after the eld title): (1) Strategy[5]. Development and timeliness of a superordinate procurement strategy, application of the superordinate procurement strategy, content and level of detail of product group strategies, application of procurement levers. (2) Organization[6]. Position of procurement within the company organizational structure, structure of the procurement department, interaction with other divisions in the company, company-wide coordination of procurement activities, interface to suppliers and supplier quality management (SQM), organizational changes. (3) Processes[7]. Early integration of procurement and supplier quality, order processes, logistics processes, supply security, and make or buy decisions. (4) Methods and tools[8]. Information management, e-procurement. (5) Human resources (HRs)[9]. Setting and controlling of targets, employee level of education, employee development and level of satisfaction, and internationality. (6) Supplier management[10]. Supplier portfolio, supplier selection, supplier controlling, supplier development, and supplier integration. These sections are analyzed and statistically evaluated in greater detail. Open-ended questions in this segment of the questionnaire supplement the data collection process. The goal of evaluating the questionnaire is to distil signicant cause-effect relationships between the answers given to these elds and overall best practice in procurement.

Before starting the evaluation, two methodical aspects still need to be mentioned: the grading of the questions and the assignment of the answers to the elds. During the preparation of the questionnaire most of the questions were designed with specic answer choices to provide a clear grading within a Likert scale with marks from 1 to 5 (Likert, 1932). Let us take, for instance, the following closed question:
To what extent is essential order information sufciently specied by internal users? (Exchange of order information with users): (1) There are no specications. (3) We get some specications but have to verify/supplement them. (5) We receive all required specications.

Benchmarking procurement functions 9

This system provides enough exibility to specically classify the answers of each BM partner. To ensure an unbiased approach to the questions, the nal version of the questionnaire sent in advance to the BM partners before the actual interviews took place contained only a fraction of these choice possibilities. As the interviews were carried out, the interview teams classied the answers of the BM partner according to this choice system. Besides, these closed questions, the questionnaire contains a portfolio of open questions for which no pre-dened set of answers were provided, for instance:
What is the inventory turnover rate in your company?

After all benchmark interviews had been carried out, the team compared all given answers in order to constitute a social basis of comparison. This process ensured an unbiased a posteriori grading of the answers. To keep the number of questions in the questionnaire to a necessary minimum and also consider all relevant inuences a single question has, the evaluation respects a multiple contribution of questions to different sectors. Hence, the questionnaire becomes like a cobweb of interlocking questions. Just take the above-mentioned closed question on order specications as an example: this question with its answer scheme located in the organization sub-eld order processes also provides insight in how the company-wide coordination of procurement activities is performed and thus contributes to a sub-eld of organization, too. It seems reasonable to present the pathway from the questionnaire to the reports, in order to have a blue print handy for other BM projects. After the questionnaire is completed in the BM interviews, the results are transferred to an Excel document which allows a digitalization of the data and a rst statistical pre-analysis[11]. Next, Excel-Macros export the digitalized data into the powerful statistics software R[12, 13]. By the use of R the statistical key measures (mean, median, quartiles, etc.) as well as the linear models, regression analysis and general studies on the dependency of the data sets that were carried out in this study were generated. In order to automatically generate the nal BM reports, the typesetting software LaTeX was used[14]. Here, one nal report is generated from which the individual reports for the BM partners can be easily separated. Furthermore, copy-and-paste errors can be avoided and the total amount of time spend on the report is reduced by not handling the data manually. 3.2 Indicators for procurement best practice Dening reliable/tangible measures for outstanding performance in complex economic situations is a tough topic (Kaplan and Norton, 1996). We choose the so-called

BIJ 17,1

10

procurement levers as these performance indicators. The procurement levers are a set/toolbox of different methods to strategically classify procurement activities. Thus, their degree or implementation serves as a good guideline of how successful the overall procurement function should be. On the basis of A.T. Kearney (2008), we asked each BM partner on how successfully on a scale from one to ve[15] the following 12 procurement levers, grouped by three themes, are applied within his/her function: (1) Commercial levers: . Pooling (bundle between different factories and use economies of scale). . Negotiation concepts (lead negotiations, follow a specic methodology, use of e-procurement). . Global sourcing (use request for information/request for quotation (often in combination with a request for proposal), optimize sourcing process, transfer volume to emerging procurement markets). . Supplier portfolio (introduce controlling tools for purchasing activities and savings, focus on core suppliers). . Target costing (break down the costs, view the life cycle costs, and total cost of ownership, make or buy decision). (2) Technical levers: . Supplier development (reduce waste, develop optimization approaches at supplier sites, supplier risk management). . Standardization (set up cross-functional teams, eliminate over-variety). . Redesign to cost (conduct function and value analyses, redesign the specication of the product). . Simplifying technical specications (reduce over-specications, implement standards, dene functional specications). (3) Supply chain process levers: . Supply chain integration (optimize material ow, warehousing, procurement systems, implement IT-solutions). . Procurement processes (accelerate the order process, standardize procurement process, long-term procurement, simultaneous procurement (2nd source)). . Supplier value integration (decide the level of outsourced process steps, cooperate and integrate suppliers). Figure 1 shows an estimation of todays degree of application of these levers. Traditionally, just the commercial levers are exploited to a considerable amount whereas technical and supply chain process levers are underdeveloped. Interviewing procurement managers, the application of these later types need to be pushed forward in order to maintain company-wide growth and not to get lost economically hard times. Comparing the 12 levers and their dening characteristics with the lists given in Table I we immediately recognize a huge overlap. Hence, taking the average over all degrees of implementation of these levers gives us a reliable number for the procurement performance (in accordance with earlier studies, Table I). To identify

the factors having a major contribution to the thus dened procurement performance, we inspect the correlation of this use of procurement levers with other variables gained from our raw data. In particular, the correlations found will (Figure 2), in a sense, enhance the degree of implementation of all levers simultaneously. 4. Empirical results and analysis Since the successful application of Gauss in 1801 linear regression/the least square method is a popular tool to t empirical data to linear laws. In breve, given empirical
3 types of levers Commercial (traditional levers) Target costing Simplifying technical specs Pooling Negotiation concepts Global sourcing

Benchmarking procurement functions 11

Supply chain process levers

Redesign/ design to cost

Supplier portfolio

Technical levers

Standardization Supplier development

Supply chain integration Procurement process Supplier value integration

Lever estimation Future target estimation

Figure 1. Traditional/todays degree of application of the 12 procurement levers (inner area within the spider diagram) and trend prognosis on how the application of these levers will have to change to ensure further growth

Standing and cooperation with in the company Cross-functional teams Hierarchical positioning within the company HR training and development Procurement unit/function Success = Application of proc. levers Expertise of the employees

Cooperation with other units/functions

Continuous evaluation of the suppliers

Supplier integration Supplier as partner

Note: Positive correlations among different key features of the questionnaire are represented by arrows

Figure 2. Dependencies

BIJ 17,1

12

variables X1, X2, . . . one can construct a linear model Y a b1 X1 e or a multi-linear model Y a b1 X1 b2 X2 e with constant coefcients a, b1, b2, . . . that governs the empirical realization. The condition under which this method can be applied to gain such correlation insights is that the error e which measures the deviation of the empirical data from the linear model is distributed normally. Figure 2 shows the key results of this section: the correlations between different features of the BM questionnaire. For instance, there is a high correlation between the use of procurement levers and integration of the procurement function within the company. In other words, this states, that the higher the integration of the procurement function is, the better is the degree of procurement lever application (Finding 1). The remaining correlations are summarized in the Findings 2-7. Though this would be enough to satisfy the requirements on any executive level, let us go one step back and discuss the evaluation of the raw data. The goal of this paper is to derive these cumulative ndings step by step, despite the fact that, traditionally, the data basis to apply advanced statistical methods is rather small. The structure of this section is aligned with this quest for major correlations: rst the use of procurement levers is examined, then that of cross-functional teams. The inuence of changes of the procurement organization could not be decided, despite the fact that all top performing companies performed major changes of their procurement organizations during the last two years[16]. Finally, the consistency of the questionnaire is checked. As the number of observations is rather small we have to consider the inuence of any outlying observations on the results of the linear model t. In order to check the assumption made by tting the multiple regression model estimated residuals are used[17]. The simplest and most useful possibility of checking these residuals is a normal probability plot of these ordered residuals. This so-called quantile-quantile (q-q) plot is a graphical technique for determining if two data sets come from populations with a common distribution and the normal probability plot assesses, whether or not a data set is approximately normally distributed (Fahrmeir et al., 2004, p. 490). This is an essential part of the mathematical methodology, as without a normal distribution of the data, the whole regression analysis is worthless. 4.1 Effects on the use of procurement levers First, the implications on the optimal use of procurement levers are analyzed by a linear regression ansatz[18] (Table II). The correlations[19] show a direct relationship between the use of procurement levers and the elds strategy (correlation . 0.60), organization, processes, methods and tools, HRs, and supplier management (each with correlation . 0.40). Note, that the data basis is too small to consider any correlation , 0.4.
Procurement levers Correlation Strategy Organization Processes Methods and tools HRs Supplier management 0.65 0.42 0.46 0.46 0.42 0.56

Comment Signicant Moderately signicant Moderately signicant Moderately signicant Moderately signicant Signicant

Table II. Effects of the use of procurement levers

During our analysis, we found that 0.75 is the highest (meaningful) correlation factor achieved, and thus categorized correlation factors between 0.4 and 0.5 as moderately signicant, factors between 0.5 and 0.7 as signicant and those above 0.7 as highly signicant (Sachs, 2002, p. 536), on condence regions for correlation coefcients. To further study the inuence on the use of the procurement levers, we separately consider their relationship to the single sub-elds. This is especially necessary as the use of procurement levers is also contained in the strategy sub eld application of procurement levers such that a highly signicant correlation trivially exists between these factors of inuence. 4.1.1 Strategy against procurement levers. The most important nding is that if you do not have a procurement strategy, the application of procurement levers is almost of no signicant impact. In case, a company does not take the time to think about a procurement strategy (e.g. number of sourcing activities in emerging procurement markets like China and India) applying the procurement lever negotiation tactics may turn out completely useless because of a lack of alternative cheap supply sources. In practical reality, before starting to use a set of levers, design a strategy and follow a roadmap. In Table III, the correlation between use of procurement levers and the sub-elds is displayed. As mentioned before, the use of procurement levers is directly included in the strategy sub eld application of procurement levers. So a correlation coefcient of 0.98 is not surprising and does not provide any new information. This phenomenon of a correlation caused by internal factors (e.g. contributing to both examined data sets) is well known in statistical analysis and called causal correlation (Sachs, 2002, p. 508), for a (pretty nice) thorough discussion of different types of causes for correlations. The remaining categories in the sub-eld strategy show that there is no relationship of signicance. This nding is rather surprising: we would expect that a stringent strategy should be the cornerstone of superior performance. Though we can argue with the difference between theory and actual living of a theory, i.e. that there has always been a gap between written strategy and day-to-day practice. In fact, the careful analysis and development of a cohesive procurement strategy and its reection in a company-wide document should be studied in further research. 4.1.2 Organization against procurement levers. Let us have a more detailed look at the relationship between organization and use of procurement levers, to interpret Table II more precisely: Table IV shows, that the most signicant aspects to enhance the use of procurement levers by procurement organization are integration of the procurement function within the company organization and interaction of the
Procurement levers Correlation Strategy development Strategy implementation Strategy commodities Application of procurement levers 0.19 0.26 0.20 0.56

Benchmarking procurement functions 13

Comment Not signicant Not signicant Not signicant Highly signicant Table III. Strategy against procurement levers

BIJ 17,1

14

procurement functions with the other units. Apparently, there is no effect by the structure of the procurement organization, coordination with other units and SQM: . Finding 1 (Procurement success depends on integration). The better the integration of the procurement unit within the company, the better is the overall application of the procurement levers and vice versa (Figure 3). . Finding 2 (Procurement success depends on cross-functional interaction). The better the cross-functional interaction of the procurement with other units, the better is the overall application of the procurement levers and vice versa. These ndings are easy to understand if we think about a realistic scenario of un-coordinated sourcing activities of isolated procurement department that function mostly as fulllment department for engineering or production. Demoted to order fulllment, not integrated into decision-making processes and not respected cross-functionally for their expertise, a lot of procurement effort just evaporates, regardless which levers are applied. Note that in Section 4.4, a correlation between integration and interaction is proven. Considering the clear result of Schuh et al. (2007) (Table I), that procurement best practice is characterized by central coordination vs local execution, we found a more informal characteristic of integration and interaction. It is worth to note, that all of the
Procurement levers Correlation

Comment Highly signicant Not signicant Signicant Not signicant Not signicant

Table IV. Details for the analysis of organization against procurement levers

Integration Structure Interactions Coordination SQM

0.75 0.1 0.6 0.29 0.04

Integration / procurement levers 4.0 Procurement levers 3.5 3.0 2.5 2.0 3.5 4.0 4.5 Integration 5.0 Residuals 0.2 0.0 0.2 0.4 0.6 0.4

Residuals Q-Q-plot Integration / procurement levers

Figure 3. Use of procurement levers against integration of procurement organization within the company

1 0 1 Theoretical quantiles

Note: The correlation of "integration" vs "procurement levers" is 0.75

companies interviewed already structured their procurement organization in the way Schuh et al. (2007) suggests. Thus, a regression ansatz, which always compares relative contexts, cannot gain any further insights (we just unable to compare different organizational concepts if there are not any). 4.1.3 HRs against procurement levers. Table V and Figure 4 allow an in-depth look at the relationship between HRs (especially education[20] and training[21]) and procurement levers, see Fawcett et al. (2004) for a broader benchmark on the efciency criteria of employees. We can state that the use of procurement levers is correlated with the training of the employees whereas the education level does neither provide a barrier nor an extra chance for the optimal exploitation of procurement levers: . Finding 3. The better the training, the better is the overall application of the procurement levers and vice versa. This nding(s) surprised us momentarily. Knowing that a lot of procurement departments of even global players and cutting edge technology companies suffer from lack of expertise, the not existing correlation between education and the (successful) use of some procurement levers might be doubted. A tough negotiator does not necessarily has to carry a MBA or PhD degree but more advanced procurement levers like reverse engineering or design to cost do very well require a more basic understanding of engineering and cost calculation. Answering this section of the questionnaire, interviewees may have been biased through their own biography. 4.1.4 Supplier management against procurement levers. One of the core competencies of the procurement function is supplier management (Youssef and Zairi, 1996a, b; Briscoe et al., 2004). It does not surprise, that a high correlation to the use of procurement levers can be found with this eld (Table VI).

Benchmarking procurement functions 15

Procurement levers Correlation Education Training 2 0.22 0.5

Comment Not signicant Moderately signicant

Table V. Use of procurement levers against education and training

Training / procurement levers 4.0 1.0 Procurement levers 3.5 3.0 2.5 2.0 1 2 3 Training 4 5 Residuals 0.5

Residuals Q-Q-plot Training / procurement levers

0.0

0.5 1 0 1 Theoretical quantiles

Note: The correlation of "training" vs "procurement levers" is 0.5

Figure 4. Use of procurement levers against training

BIJ 17,1

16

It is important to note, that the single sub-eld supplier integration correlates rather strongly with the overall use of (all) procurement levers (commercial, technical, SCM) (Figure 5): . Finding 4 (Procurement success depends on supplier integration). The better the supplier integration, the better is the overall application of the procurement levers and vice versa. . Finding 5 (Procurement success depends on supplier evaluation). The better the supplier evaluation process (controlling), the better is the overall application of the procurement levers and vice versa. The above-mentioned ndings are pretty much self-explaining: a high degree of integration does denitely facilitate the application of almost every procurement lever. It is much easier to get into price negotiations or reverse engineering projects with suppliers closely aligned than others only remotely coordinated. Not to mention a faster pace of getting results, a higher product-to-market-rate and reduced failure rate. Also the correlation of procurement success and supplier evaluation does not come as a complete surprise. The more time you spend on pre-screening and ltering the set of possible suppliers, the more professional you calibrate lter criteria and the better you integrate an expert team of relevant departments, the higher the quality of the selected base of suppliers.
Procurement levers Correlation Supplier Supplier Supplier Supplier Supplier SQM portfolio selection controlling development integration 0.35 2 0.04 0.53 0.5 0.68 0.04

Comment Not signicant Not signicant Signicant Moderately signicant Signicant Not signicant

Table VI. Use of procurement levers against supplier integration

SPM (integration) / procurement levers 4.0 3.5 3.0 2.5 2.0 Residuals 0.6 0.4 0.2 0.0 0.2 0.4

Residuals Q-Q-plot SPM (integration) / procurement levers

Procurement levers

Figure 5. Use of procurement levers against supplier integration

0.6 1 2 3 SPM (integration) 4 5 1 0 1 Theoretical quantiles

Note: The correlation of "supplier portfolio management (integration)" vs "procurement levers" is 0.68

Comparing our results with Section 2/Table I, we recognize that the supplier portfolio and supplier selection as a key driver for procurement performance could not be rediscovered. Again, we seem to have a discrepancy between best possible theoretical procurement strategies and actual living of these strategies. 4.2 Cross-functional teams against HRs Second, the inuence factors on cross-functional teams are inquired: the most signicant correlation is that to HRs (Table VII and Figure 6). The single most important factor for the correlation with cross-functional teams is training: Figure 6 shows a relationship of signicance between the training[21] of staff and the integration of cross-functional teams. It is quite interesting to note, that the sub-elds for training,, i.e. existence of a training plan, career advancement, and the measure of the employees satisfaction, alone do not lead to high-correlation factors, but instead enhance each other positively to gain the high-correlation factor of training: . Finding 6 (A holistic staff development program is key for cross-functional teams). The better the training, the better is the efcient cooperation of cross-functional teams and vice versa (Figure 7).
Cross-functional teams Correlation Education Language competence Level of qualications Level of special education Training Training plan Career advancement Measurement of employee satisfaction Internationality 0.2 0.01 0.01 0.37 0.6 0.37 0.49 0.46 0.1

Benchmarking procurement functions 17

Comment Not signicant Not signicant Not signicant Not signicant Signicant Not signicant Moderately signicant Moderately signicant Not signicant

Table VII. HRs against cross-functional teams

HR / cross-functional teams 0.6 Crossfunctional teams 4.0 Residuals 0.4 0.2 0.0 0.2 0.4 2.5 0.6 2.5 3.0 3.5 4.0

Residuals Q-Q-plot HR / cross-functional teams

3.5

3.0

Human resources

0 1 Theoretical quantiles

Figure 6. HRs against cross-functional teams

Note: The correlation is 0.5

BIJ 17,1

Moreover, signicant correlations between cross-functional teams and integration or interactions can be detected. There is a relationship highly signicant relationship of cross-functional teams and procurement levers, too. The corresponding correlations are summarized in Table VIII: . Finding 7: The better the integration (organization), the better is the efcient cooperation of cross-functional teams and vice versa. The better the interactions (organization), the better is the efcient cooperation of cross-functional teams and vice versa. The better the overall application of the procurement levers, the better is the efcient cooperation of cross-functional teams and vice versa. This section provides some answers to questions regarding the impact of motivation, incentive, career opportunities and the overall appreciation of working in the procurement department within a company (i.e. cross-functional teams). Still, the procurement department does not belong to the chosen few departments where fast track careers are developed. Sales and marketing, production, research departments are considered better places to start a successful company career and learn the trade.
Residuals Q-Q-plot HR (training) / cross-functional teams 0.6 Cross-functional teams 4.0 0.4 0.2 Residuals 1 2 3 4 Human resources (training) 5 0.0 0.2 0.4 0.6

18

HR (training) / cross-functional teams

3.5

3.0

2.5

Figure 7. Training against cross-functional teams

0 1 Theoretical quantiles

Note: The correlation is 0.6

Cross-functional teams Correlation Integration Interactions Procurement levers Table VIII. 0.64 0.76 0.64

Comment Signicant Highly signicant Signicant

Note: Correlations between organization details (integration and interactions), procurement levers, and cross-functional teams

Therefore, training facilities and career advancement plans are a vital part for successfully integrate procurement staff into cross-functional teams and leverage procurement levers. 4.3 Analysis of effects by organizational changes Finally, we ask whether there is any relation between organizational changes and the six elds (strategy, organization, processes, methods and tools, HRs, and supplier management). Table IX shows the corresponding correlations: apparently the organizational changes of the last two years in the procurement department have no detectable effect on strategy, organization, processes, HRs, and supplier management (correlation , 0.10). As you can see in Table IX the correlation between organizational changes and methods and tools is 0.22. Thus, if there are any effects by organizational changes they show up in methods (information management) and tools for e-procurement. (Note that the data basis is too small to consider any correlation , 0.4). 4.4 Internal correlation analysis Last, we check the consistency of the given answers of the BM partners. This is done in a two-step approach: (1) We know that there are specic correlations within the sub-elds. (2) We check by means of correlation matrices if these relations occur in the answers. To our rst item: the questionnaire has specic correlations between some of the sub-elds of one eld, in particular, we assume: . strategy development against strategy implementation; . integration of organization against interactions; . coordination against interactions; . process security against involvement; . ordering processes against logistic processes; . supplier portfolio against supplier controlling; . supplier portfolio against supplier development; . supplier portfolio against supplier integration; and . supplier development against supplier integration.

Benchmarking procurement functions 19

Organizational change Correlation Strategy Organization Processes Methods and tools HRs Supplier management 0.03 20.06 0.05 0.22 20.03 20.09

Comment Not Not Not Not Not Not signicant signicant signicant signicant signicant signicant

Table IX. Effects by organizational changes

BIJ 17,1

20

Next, Tables X-XV show the correlation matrices for the single elds. Here, we see exactly the expected consistent behavior of the answers, despite internationality and targets at the eld HRs. We assume this side effect to be due to the small amount of gathered data. A careful inquiry of internationality and targets shows a rather instable behavior of the regression line with clearly patterns of the underlying data (column structure and staircase behavior of residual plot (Figure 8). Another ansatz would be to design the questionnaire a prior in such a way, that no interdependencies between the sub-elds are expected and later on check in the just conducted way, whether the correlation matrices are unit matrices (like in Table XIII). Thus, we really have a consistent set of answers. This fact is supported by the impressions of the interview teams, that none of the BM partners was holding back information or construction unreliable positive statements about his procurement function. sume 5. Re Throughout this paper, we have set-up a metric to measure procurement best practice by means of day-to-day tasks to be accomplished in order to select best suppliers and to implement cost cutting activities: the procurement levers. This metric was applied to reevaluate the performance indicators derived in earlier studies (as shown in Section 2).
Development Implementation 0.66 1 20.33 0.26 Commodity strategy 20.16 20.33 1 0.2 Procurement levers 0.19 0.26 0.2 1

Table X. Internal correlation strategy

Development Implementation Commodity strategy Procurement levers

1 0.66 20.16 0.19

Integration Integration Structure Interactions Coordination SQM 1 0.03 0.68 0.48 0.12

Structure 0.03 1 0.28 0.01 0.29

Interactions 0.68 0.28 1 0.62 0.34

Coordination 0.48 0.01 0.62 1 0.39

SQM 0.12 0.29 0.34 0.39 1

Table XI. Internal correlation organization

Involvement Table XII. Internal correlation processes Involvement Ordering Logistics Security 1 0.28 0.44 0.53

Ordering 0.28 1 0.53 20.03

Logistics 0.44 0.53 1 0.09

Security 0.53 2 0.03 0.09 1

Information management Table XIII. Internal correlation methods and tools Information management e-Procurement 1 2 0.02

e-Procurement 20.02 1

The results of the regression analysis were further claried and condensed into six key characteristics of superior performance (Figure 2). Table XVI compares our nding to the studies cited in Section 2 and displays the overlap between the known indicators and our results. Though some intuitively plausible indicators for best practice identied in earlier studies could not been rediscovered. In particular, we mentioned the coherent formalization of a holistic procurement strategy and the structuring of the supplier portfolio. Concerning the still relative small data basis, these items should be claried in further studies. Moreover, practical guidelines should be establishes to put these and further ndings into actual procurement day-to-day practice.

Benchmarking procurement functions 21

Targets Targets Education Training Internationality 1 20.13 0.37 0.52

Education 20.13 1 0.04 0.12

Training 0.37 0.04 1 0.44

Internationality 0.52 0.12 0.44 1

Table XIV. Internal correlation HRs

Supplier portfolio Portfolio Selection Controlling Development Integration SQM 1 0.11 0.57 0.53 0.52 0.05

Supplier selection 0.11 1 0.24 0.05 20.15 0.24

Supplier controlling 0.57 0.24 1 0.49 0.22 0.11

Supplier development 0.53 0.05 0.49 1 0.63 0.32

Supplier integration 0.52 20.15 0.22 0.63 1 0.06

SQM 0.05 0.24 0.11 0.32 0.06 1

Table XV. Internal correlation supplier management

HR (targets)/ internationality 5.0 4.5 Internationality 4.0 3.5 3.0 2.5 1.5 2.0 2.5 3.0 3.5 HR (targets) 4.0 4.5 1.0 Residuals 0.5

Residuals Q-Q-plot HR (targets)/ internationality

0.0

0.5

1 0 1 Theoretical quantiles

Figure 8. Internationality against targets

BIJ 17,1

Field of excellence

Indicator (as found in our study)

Comparison with other studies (Table I) Senior management support for procurement Corporate thinking and crossfunctional responsibility Highly qualied buyers Procurement personnel must be on face value with other units Holistic supplier evaluation Supplier value integration and early involvement of key suppliers

22
Table XVI. Comparison of our ndings with the results provided in Table I

Strategy and cross-company Hierarchical positioning within coordination the company Cooperation with other units/ functions Cross-functional teams HRs HR training and development Supplier management Continuous evaluation of the suppliers Supplier integration

Notes 1. See Brookshaw and Terziovski (1997) for the importance of procurement with respect to quality management instead of a mere cost perspective. 2. Because of the relative small number of BM partners and the fact that some of them are direct competitors (not to our clients company but among each other), the names of the benchmark partners need to be kept anonymous. 3. See, in particular Hyland and Beckett (2002) for the value of internal BM. 4. Tough, the answer alternatives for closed questions given in the questionnaire were selected with greatest care, often times the situation at the BM partners differed and the BM partners made mental leaps to forthcoming questions or provided additional valuable information. 5. Inter alias, the following sources were used to design the questions in this eld: Frehner and Bodmer (2000), Hahn and Kaufmann (2000), Large (2006), Schneider (1990) and Stark (1994). 6. Inter alias, the following sources were used to design the questions in this eld: Corey (1978). 7. Inter alias, the following sources were used to design the questions in this eld: Boutellier et al. (2002), Strache (1991) and Wagner and Weber (2006). 8. Inter alias, the following sources were used to design the questions in this eld: Kerkhoff (2006), Nekolar (2002) and Nepelski (2006). 9. Inter alias, the following sources were used to design the questions in this eld: Frohlich-Glantschnig (2005). 10. Inter alias, the following sources were used to design the questions in this eld: Ferreras (2007), Hartmann et al. (2004), Janker (2004), Schiele (2006), Jahns and Moser (2005) and Jahns and Moser (2006). 11. See, e.g. Hofmann and May (1999) for an introduction into statistical analysis with Excel. 12. R is a language and environment for statistical computing and graphics. It is a GNU project which is similar to the S language and environment that was developed at Bell Laboratories (formerly AT&T, now Lucent Technologies) by John Chambers and colleagues. R can be considered as a different implementation of S. One of Rs strengths is the ease with which well-designed publication-quality plots can be produced. Great care has been taken over the defaults for the minor design choices in graphics (available at: www.r-project.org/). 13. For references on statistics with R (Becker and Chambers, 1986; Dolic, 2003; Maindonald and Braun, 2003; Murrell, 2005; Everitt and Hothorn, 2006; Ligges, 2006).

14. LaTeX offers programmable desktop publishing features and extensive facilities for automating most aspects of typesetting and desktop publishing, including numbering and cross-referencing, tables and gures, page layout and bibliographies. 15. 1 was considered as practically no use of the specic lever and 5 as its total exploitation, i.e. the company does not see any way to further increase the use of this lever. The advantage of taking the procurement levers as a basis for further research is, that the degree of application of some of them can be directly gained from procurement data bases and an easy to conduct survey can be established among the buyers on what percentage of their contracts/commodities which lever was used during a specic time interval. 16. It seems that all these changes aimed to enhance the important elds of integration, interaction, and cross-functional teams to promote/enable procurement success. 17. The application of statistics and regression analysis in business and engineering has a long and fruitful history (Hald, 1952; Dielman, 1996; Czitrom and Spagon, 1997). 18. For an introducing text on linear regression (Fahrmeir et al., 2004, p. 476; Assenmacher, 2003, p. 182; Maindonald and Braun, 2003, p. 107). 19. For an introducing text on correlation analysis (Kleinbaum and Kupper, 1978, p. 71; Sachs, 2002, p. 493). 20. That is to say, language competence, level of qualications, and level of education. 21. That is to say, employee development and level of satisfaction: existence of training/continuing education plan, promotion of the development potential of employees, measurement of the level of employee satisfaction. References Aberdeen Group (2006), Global Supply Chain Benchmark Report. Industry Priorities for Visibility, B2B Collaboration, Trade Compliance, and Risk Management, Aberdeen Group, Boston, MA. Andersen, B., Fagerhaug, T., Randml, S., Schuldmaier, J. and Prenninger, J. (1990), Benchmarking supply chain management: nding best practices, Journal of Business & Industrial Marketing, Vol. 14 Nos 5/6, pp. 378-89. Assenmacher, W. (2003), Deskriptive Statistik, 3rd ed., Springer, Berlin. A.T. Kearney (2004), Creating Value through Strategic Supply Chain Management 2004 Assessment of Excellence in Procurement, A.T. Kearney, Marketing & Communications, Chicago, IL. A.T. Kearney (2008), The Purchasing Chessboard Buying in Turbulent Times, A.T. Kearney, Marketing & Communications, London. Becker, R. and Chambers, J. (1986), The New S Language: A Programming Environment for Data Analysis and Graphics, Wadsworth & Brooks/Cole Advanced Books, Monterey, CA. Boutellier, R., Gassmann, O. and Voit, E. (2002), Projektmanagement in der Beschaffung. Zusammenarbeit von Einkauf und Entwicklung, 2nd ed., Hanser Wirtschaft, Munchen. Brandmeier, R., Hofmann, T. and Rupp, F. (2008), Erfahrungen bei der Kombination von Strategie- und Markt-Benchmarking im Einkauf, PPS-Management, No. 4, pp. 53-6. Briscoe, J., Lee, T. and Fawcett, S. (2004), Benchmarking challenges to supply-chain integration: managing quality upstream in the semiconductor industry, Benchmarking: An International Journal, Vol. 11 No. 2, pp. 143-55. Brookshaw, T. and Terziovski, M. (1997), The relationship between strategic purchasing and customer satisfaction within a total quality management environment, Benchmarking for Quality Management & Technology, Vol. 4 No. 4, pp. 244-58.

Benchmarking procurement functions 23

BIJ 17,1

24

Camp, R. (1989), Benchmarking: The Search for Industry Best Practices That Lead to Superior Performance, ASQC Quality Press, Milwaukee, WI. Carpinetti, L. and de Melo, A. (2002), What to benchmark? A systematic approach and cases, Benchmarking: An International Journal, Vol. 9 No. 3, pp. 244-55. Codling, B. (1997), Dynamics of best practice a multidimensional perspective, Benchmarking for Quality Management & Technology, Vol. 4 No. 2, pp. 69-103. Codling, B. (1998), Benchgrafting: a model for successful implementation of the conclusions of benchmarking studies, Benchmarking for Quality Management & Technology, Vol. 5 No. 3, pp. 158-64. Corey, R. (1978), Procurement Management: Strategy, Organization and Decision-Making, CBI, Boston, MA. Czitrom, V. and Spagon, P. (1997), Statistical Case Studies for Industrial Process Improvement, ASA-SIAM Series on Statistical and Applied Probability, ASA-SIAM, Alexandria, VA. Dielman, T. (1996), Applied Regression Analysis for Business and Economics, 2nd ed., Duxbury Press, Belmont, CA. Dolic, D. (2003), Statistik mit R. Einfuhrung fur Wirtschafts- und Sozialwissenschaftler, Oldenbourg, Munchen. Everitt, B. and Hothorn, T. (2006), Statistical Analysis Using R, Chapman & Hall/CRC, Cambridge. Fahrmeir, L., Kunstler, R. and Pigeot, I. (2004), Statistik, 5th ed., Springer, Berlin. Fawcett, S., Rhoads, G. and Burnah, P. (2004), People as the bridge to competitiveness: benchmarking the ABCs of an empowered workforce, Benchmarking: An International Journal, Vol. 11 No. 4, pp. 346-60. Ferreras, M. (2007), Beziehungsmanagement zu Lieferanten. Nachhaltiger Unternehmenserfolg durch neue Wege der Lieferantenbewertung, Vdm Verlag Dr. Muller, Saarbrucken. Frehner, U. and Bodmer, C. (2000), Best Practice im Einkauf. Optimieren durch messen und vergleichen, Fachbuchverlag Leipzig, Leipzig. Frohlich-Glantschnig, E. (2005), Berufsbilder in der Beschaffung, Ergebnisse einer Delphi-Studie, Deutscher Universitatsverlag, Wiesbaden. Gilmour, P. (1998), Benchmarking supply chain operations, Benchmarking for Quality Management & Technology, Vol. 5 No. 4, pp. 283-90. Hahn, D. and Kaufmann, L. (2000), Handbuch Industrielles Beschaffungsmanagement, Gabler, Wiesbaden. Hald, A. (1952), Statistical Theory with Engineering Applications, Wiley, New York, NY. Hartmann, H., Orths, H. and Pahl, H.-J. (2004), Lieferantenbewertung, aber wie?, 3rd ed., Deutscher Betriebswirte, Gernsbach. Hofmann, J. and May, S. (1999), Anwendungsorientierte Statistik mit Excel Band 1. Deskriptive Statistik und Wahrscheinlichkeitstheorie, Oldenbourg, Munchen. tztes Benchmarking Homburg, C., Werner, H. and Englisch, M. (1997), Kennzahlengestu im Beschaffungsbereich: Konzeptionelle Aspekte und empirische Befunde, Die Betriebswirtschaftslehre, Vol. 57, pp. 48-64. Hyland, P. and Beckett, R. (2002), Learning to compete: the value of internal benchmarking, Benchmarking: An International Journal, Vol. 9 No. 3, pp. 293-304. Jahns, C. and Moser, R. (2005), Supplier Value. How to Dene and Integrate the Value of Your Suppliers, Verlag Wissenschaft & Praxis, St Gallen.

Jahns, C. and Moser, R. (2006), Schaffung von Wettbewerbsvorteilen durch die Auswahl strategischer Lieferanten, European Business School (EBS), SMI (Supply Management Institute), International University Bad Reichartshausen, Wiesbaden. Janker, C.G. (2004), Multivariate Lieferantenbewertung, Deutscher Universitatsverlag, Wiesbaden. Kaplan, R. and Norton, D. (1996), The Balanced Scorecard, Harvard Business School Press, Boston, MA. Kerkhoff, G. (2006), Global Sourcing, Wiley-VCH, Weinheim. Kleinbaum, D.G. and Kupper, L.L. (1978), Applied Regression Analysis and Other Multivariable Methods, Duxbury Press, Belmont, CA. Large, R. (2006), Strategisches Beschaffungsmanagement. Eine praxisorientierte Einfuhrung, 3rd ed., Gabler, Wiesbaden. Le Sueur, M. and Dale, B. (1997), Benchmarking: a study in the supply and distribution of spare parts in a utility, Benchmarking for Quality Management & Technology, Vol. 4 No. 3, pp. 189-201. Ligges, U. (2006), Programmieren mit R, 2nd ed., Springer, Berlin. Likert, R. (1932), A technique for the measurement of attitudes, Archives of Psychology, Vol. 140, pp. 1-55. Maindonald, J. and Braun, J. (2003), Data Analysis and Graphics Using R An Example-Based Approach, Cambridge University Press, Cambridge. Murrell, P. (2005), R Graphics (Computer Science and Data Analysis), Chapman & Hall/CRC, Cambridge. Nekolar, A.-P. (2002), e-Procurement, Springer, Berlin. Nepelski, D. (2006), The impact of e-procurement on the number of suppliers: where to move to?, Discussion Papers of DIW Berlin No. 587, DIW Berlin, German Institute for Economic Research, Berlin, pp. 1-30. Raymond, J. (2008), Benchmarking in public procurement, Benchmarking: An International Journal, Vol. 15 No. 6, pp. 782-93. Razmi, J., Zairi, M. and Jarrar, Y. (2000), The application of graphical techniques in evaluating benchmarking partners, Benchmarking: An International Journal, Vol. 7 No. 4, pp. 304-14. Saad, M. and Patel, B. (2006), An investigation of supply chain performance measurement in the Indian automotive sector, Benchmarking: An International Journal, Vol. 13 No. 1, pp. 36-53. Saad, M., Gunasekaran, A. and Koh, L. (2005), A business model for uncertainty management, Benchmarking International Journal, Vol. 12 No. 4, pp. 383-400. Sachs, L. (2002), Angewandte Statsitik Anwendung statistischer Methoden, 11th ed., Springer, Berlin. Sanchez-Rodrguez, C., Martnez-Lorente, A. and Clavel, J. (2003), Benchmarking in the purchasing function and its impact on purchasing and business performance, Benchmarking: An International Journal, Vol. 10 No. 5, pp. 457-71. Schiele, H. (2006), How to distinguish innovative suppliers? Identifying innovative suppliers as new task for procurement, Industrial Marketing Management, No. 35, pp. 925-35. Schneider, B. (1990), Einkaufsvorbereitung, Gabler, Wiesbaden. Schuh, G., Haag, C. and Moller, H. (2007), Zehn Erfolgsfaktoren oder wie der Einkauf adverse Selektion verhindern kann, Beschaffung Aktuell, Vol. 6, pp. 18-21 and Vol. 7, pp. 16-18. Stark, H. (1994), Beschaffungsplannung und Budgetierung, 4th ed., Gabler, Wiesbaden.

Benchmarking procurement functions 25

BIJ 17,1

26

Strache, H. (1991), Analyse und Bewertung von Fremd- und Eigenleistungen (Make or Buy), 2nd ed., Gabler, Wiesbaden. Wagner, S. and Weber, J. (2006), Beschaffungscontrolling. Den Wertbeitrag der Beschaffung messen und optimieren, Wiley-VCH, Weinheim. Wong, W. and Wong, K. (2008), A review on benchmarking of supply chain performance measures, Benchmarking: An International Journal, Vol. 15 No. 1, pp. 25-51. Yasin, M. (2002), The theory and practice of benchmarking: then and now, Benchmarking: An International Journal, Vol. 9 No. 3, pp. 217-43. Youssef, M. and Zairi, M. (1996a), Benchmarking supplier partnerships in the context of advanced manufacturing technology implementation, Benchmarking for Quality Management & Technology, Vol. 3 No. 3, pp. 4-20. Youssef, M. and Zairi, M. (1996b), Supplier selection in an advanced manufacturing technology environment: an optimization model, Benchmarking for Quality Management & Technology, Vol. 3 No. 4, pp. 60-72. Corresponding author Florian Rupp can be contacted at: rupp@ma.tum.de

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com Or visit our web site for further details: www.emeraldinsight.com/reprints

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.